ScreeNOT: Exact MSE-Optimal Singular Value Thresholding in Correlated Noise

by   David L. Donoho, et al.

We derive a formula for optimal hard thresholding of the singular value decomposition in the presence of correlated additive noise; although it nominally involves unobservables, we show how to apply it even where the noise covariance structure is not a-priori known or is not independently estimable. The proposed method, which we call ScreeNOT, is a mathematically solid alternative to Cattell's ever-popular but vague Scree Plot heuristic from 1966. ScreeNOT has a surprising oracle property: it typically achieves exactly, in large finite samples, the lowest possible MSE for matrix recovery, on each given problem instance - i.e. the specific threshold it selects gives exactly the smallest achievable MSE loss among all possible threshold choices for that noisy dataset and that unknown underlying true low rank model. The method is computationally efficient and robust against perturbations of the underlying covariance structure. Our results depend on the assumption that the singular values of the noise have a limiting empirical distribution of compact support; this model, which is standard in random matrix theory, is satisfied by many models exhibiting either cross-row correlation structure or cross-column correlation structure, and also by many situations where there is inter-element correlation structure. Simulations demonstrate the effectiveness of the method even at moderate matrix sizes. The paper is supplemented by ready-to-use software packages implementing the proposed algorithm.


page 15

page 39


OptShrink: An algorithm for improved low-rank signal matrix denoising by optimal, data-driven singular value shrinkage

The truncated singular value decomposition (SVD) of the measurement matr...

Deep learned SVT: Unrolling singular value thresholding to obtain better MSE

Affine rank minimization problem is the generalized version of low rank ...

Fast Algorithm for Low-rank matrix recovery in Poisson noise

This paper describes a fast algorithm for recovering low-rank matrices f...

Optimal cleaning for singular values of cross-covariance matrices

We give a new algorithm for the estimation of the cross-covariance matri...

Optimal shrinkage of singular values under high-dimensional noise with separable covariance structure

We consider an optimal shrinkage algorithm that depends on an effective ...

An exact sinΘ formula for matrix perturbation analysis and its applications

Singular vector perturbation is an important topic in numerical analysis...

Robust Integrative Biclustering for Multi-view Data

In many biomedical research, multiple views of data (e.g., genomics, pro...

Please sign up or login with your details

Forgot password? Click here to reset